[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: overall performance
From: |
Mats Bengtsson |
Subject: |
Re: overall performance |
Date: |
Mon, 04 Dec 2006 08:35:56 +0100 |
User-agent: |
Thunderbird 1.5.0.7 (X11/20060909) |
Comparing to the -O flags of most compilers, to optimize the generated code
at the expense of longer compilation time, I see both similarities and
differences:
- It's something you turn on when you know that you input source code is
more or less correct.
- Most users don't know and don't care exactly what extra optimization
is done.
(Even if a compiler often has many different levels of optimization
available
and a list in the documentation of what each level does, I'm sure that
almost
all users choose the highest level of optimization for production code.)
* If the compiler is bug free, the resulting program should be equivalent
no matter if you use -O or not. However, in the case of LilyPond, the
semantics of the music will be the same, but not the layout.
* In the case of LilyPond, this is something you turn on before doing the
final tweaks of the layout, which means that the turn-around time for the
fine tuning part of the typesetting will be affected by the longer
compilation
time.
/Mats
Graham Percival wrote:
Would it be desirable to have
\fastProcessing
\qualityProcessing
macros?
No, this is as stupid as the 'windows video acceleration' settings:
most users won't know they're there,
Well, that's what the documentation is for.
and if they know, it's not clear what the settings do.
That's what ly/property-init.ly is for.
We occasionally get people (even me!) sending "bug" reports about
\override Score.PaperColumn #'keep-inside-line = ##t
You said that it added 5% or 10% to the processing time, so it wasn't
worth having it on by default. That's fine; I'm not criticizing that
decision. But what if somebody _is_ willing to spend that extra
processing time to get great-looking music? Currently they need to
comb through the manual (and program reference, because I haven't
documented every case because I don't know them all!) to find all such
options.
At the very least we should have a doc page with the "higher quality
output commands that are turned off by default" \override commands.
Why not also make a
\qualifyProcessing
macro that sets it up for them as well?
Cheers,
- Graham
_______________________________________________
lilypond-devel mailing list
address@hidden
http://lists.gnu.org/mailman/listinfo/lilypond-devel
--
=============================================
Mats Bengtsson
Signal Processing
Signals, Sensors and Systems
Royal Institute of Technology
SE-100 44 STOCKHOLM
Sweden
Phone: (+46) 8 790 8463
Fax: (+46) 8 790 7260
Email: address@hidden
WWW: http://www.s3.kth.se/~mabe
=============================================